Based On The Definition Of Multi-strategy Synonymous Term Identification Method Research

Through summarizing the synonyms recognition method and the synonyms recognition method based on the definition,this paper analyzes the principle,advantages and disadvantages of each method in the application process,and finds that although each method has certain advantages,but can not recognize the synonyms comprehensively.In order to make up for the deficiency of single method,a multi strategy fusion method is put forward to realize the complementary recognition of synonyms.First through the summary of the Multi Strategy fusion method,this paper analyzes principle,advantages and disadvantages of the strategy fusion method.Then through the summary the research status of multi strategy synonyms recognition method in domestic and foreign,this paper finds the majority of the fusion methods are combined with deduplication and weighted fusion,but the fusion method has not been fully utilized.So,this paper uses three methods for multi-strategy fusion experiments,and discusses the effect of the use of multi-strategy fusion method for synonymous terms recognition.This experiment set up the test set on the Chinese scientific and technical vocabulary system(new energy vehicles).First,from three angles of literal,syntactic and semantic,the literal similarity algorithm,pattern matching algorithms and definition matching algorithms are selected for synonymous terms recognition experiments.Then,the Linear weighted combination method,SVM based machine learning method and the improved multi strategy fuzzy comprehensive evaluation method are choosed from applicable angle for multi strategies fusion experiments.Finally,the precision,recall,F value is used to evaluation the experimental results.The experimental results show that the accuracy rate and recall rate of the three single synonymous terms recognition methods is low,but the coincidence rate of the results recognized by this three method are low.So,this paper uses multi strategy fusion method to integrate single methods complementarity,to recognize synonymous terms from all aspects and levels,and to improve accuracy rate and recall rate of the synonymous terms recognition.